Human Herpesvirus 8 (HHV-8), the newly characterized oncogenic gamma-herpesvirus associated with Kaposi's sarcoma, is an important cause of morbidity and mortality, particularly in HIV infection. Yet the epidemiologic aspects of HHV-8 transmission remain poorly understood. We propose a series of studies to explore the hypotheses that: (1) Sexual transmission is an important mode of HHV-8 infection; (2) Sexual transmission is inefficient, except from persons with HIV infection, as HHV-8 shedding from the genital and oral mucosa is under tight immunologic control; and (3) Acquisition of HHV-8 infection is associated with a clinical syndrome and characterized by transiently high levels of HHV8 shedding. Our specific aims are: 1) To determine the prevalence and the incidence of HHV-8 and the biologic and behavioral risk factors for acquisition of HHV-8 among high-risk cohorts: HIV- and HIV+ men who have sex with men (MSM) in Seattle and female sex workers in Nairobi, Kenya, and 2) To determine the most likely reservoir of infectious virus and the mode of transmission of HHV-8, by sampling oral and genital secretions, in various HHV-8-infected populations, and 3) To characterize the clinical and virologic manifestations of primary HHV-8 infection, by identifying and conducting detailed evaluations of incident cases. The study will include 4 longitudinal cohorts of high-risk persons who will be examined for prevalent and incident HHV-8 infection. 2 cohorts will be prospective and include MSM in Seattle. Most will be HIV- but a substantial minority will also have HIV infection. These cohorts will be followed for 3 (n=650) and 4 (n=400) years. A third cohort will involve female sex workers in the Korogocho region of Nairobi and be followed for 3 years (n-300). Interim behavioral information, blood and samples of genital and oral secretions will be collected on all participants. The 4 cohort comprises of approximately 500 men (50% HIV-, 50% HIV+) who have been followed for anal dysplasia. Blood, anal swab and biopsy samples are available. Persons who seroconvert to HHV-8 during the study will be identified prospectively and evaluated intensively for HHV-8 shedding in the genital secretions and the for HHV-8 viremia. The analysis will focus on biologic and behavioral correlations between prevalent and incident HHV-8 infections, and on predictors of HHV-8 shedding from the genital and oral mucosa.